- 5th Aug 2022
- 06:03 am
Que 1: How should you set the Rocchio parameters ?, ?, and ? depending on what type of feedback you are using? That is, should the parameters be set differently if you are using pseudo feedback compared to user-supplied relevance judgements? What about implicit feedback through clickthrough data?
Ans: Parameters ?, ?, and ? are weights that control the amount of movement of the original vector. In terms of movement, the term in the original query are boasted by the parameter ?, the term for positive documents are boasted by the parameter ?, while term for negative documents are shrunk by the parameter ?. The parameters should be set differently if you are using pseudo feedback compared to user-supplied relevance judgments. pseudo feedback does not involve the users since it is assume that the top ranked documents, while user-supplied relevance judgments does involve the users because feedback information user is based on judgments of the search results.
Que 2: Say one of your new strategies is to measure the amount of time t a user spends on each search result document. How can you incorporate this t for each document into a feedback measure for a particular query?
Ans: One way in which t can be incorporated for each document into a feedback measure for a particular query is that t can be the time it takes a user to complete a search task for a particular query. Another is that t can be used as the length of time it takes a user to engage in viewing a search result of a specified task as a result of a query. Also, t can be used to measure the index of efficiency for performance, effectiveness, and user satisfaction of a particular query from the search engine. Moreover, t can be used to measure as a user captures information to save as the first relevant result and other relevant documents that are required to save for a certain period of time. Furthermore, t can be used to measure the needed actions or steps to be followed by the user for one particular search operation. Finally, t can be used to measure the user's trend which is useful to approximate when, what, and how long the search operation is incorporated for executing a required query.
